First cycle
degree courses
Second cycle
degree courses
Single cycle
degree courses
School of Science
DATA SCIENCE
Course unit
DEEP LEARNING
SCP9087561, A.A. 2019/20

Information concerning the students who enrolled in A.Y. 2019/20

Information on the course unit
Degree course Second cycle degree in
DATA SCIENCE
SC2377, Degree course structure A.Y. 2017/18, A.Y. 2019/20
N0
bring this page
with you
Number of ECTS credits allocated 6.0
Type of assessment Mark
Course unit English denomination DEEP LEARNING
Website of the academic structure http://datascience.scienze.unipd.it/2019/laurea_magistrale
Department of reference Department of Mathematics
Mandatory attendance No
Language of instruction English
Branch PADOVA
Single Course unit The Course unit can be attended under the option Single Course unit attendance
Optional Course unit The Course unit can be chosen as Optional Course unit

Lecturers
Teacher in charge ALESSANDRO SPERDUTI INF/01
Other lecturers NICOLO' NAVARIN INF/01

Mutuating
Course unit code Course unit name Teacher in charge Degree course code
SCP9087561 DEEP LEARNING ALESSANDRO SPERDUTI SC1176

ECTS: details
Type Scientific-Disciplinary Sector Credits allocated
Core courses INF/01 Computer Science 6.0

Course unit organization
Period Second semester
Year 1st Year
Teaching method frontal

Type of hours Credits Teaching
hours
Hours of
Individual study
Shifts
Lecture 6.0 48 102.0 No turn

Calendar
Start of activities 02/03/2020
End of activities 12/06/2020
Show course schedule 2019/20 Reg.2017 course timetable

Examination board
Examination board not defined

Syllabus
Prerequisites: It is advisable to have the basic knowledge related to Probability, Programming, and Algorithms.
Target skills and knowledge: The course covers the basic concepts related to Deep Learning, i.e. machine learning through neural networks. The mathematical concepts needed for a full understanding of the subject will be recalled. Deep feedforward neural networks will be discussed as well as the related regularization and optimization techniques for training deep models.
Basic concepts about convolutional neural networks will be introduced. As for the modelling of sequences, recurrent neural networks will be presented, with specific emphasis on the use of LSTM and similar units. Finally, we will deal with autoencoders and deep generative models. Furthermore, the TensorFlow platform will be introduced for the implementation of the models presented in the course,
Examination methods: The student must pass a written exam. In addition, the student must develop a notebook agreed with the teacher.
Assessment criteria: The student's evaluation is based on an assessment of learning of the basic concepts introduced during the course and on the analytical skills of the student.
The evaluation of the notebook considers the student's ability to identify an appropriate case study and to independently carry out a qualitatively appropriate design and implementation activity.
Course unit contents: The topics covered in the course are as follows:
- Introduction to the course contents;
- Deep Feedforward Networks;
- Regularization for Deep Learning;
- Optimization for training Deep Models;
- Basic concepts for Convolutional Neural Networks;
- Recurrent Neural Networks for sequence modelling;
- Autoencoder
- Deep Generative Models;
- TensorFlow.
Planned learning activities and teaching methods: The course will have frontal lessons.
Additional notes about suggested reading: Additional material will be available on the course website.
Textbooks (and optional supplementary readings)
  • Goodfellow, Ian; Courville, Aaron, Deep learningIan Goodfellow, Yoshua Bengio and Aaron Courville. Cambridge: MA [etc.], MIT Press, 2016. Cerca nel catalogo

Innovative teaching methods: Teaching and learning strategies
  • Lecturing

Innovative teaching methods: Software or applications used
  • Moodle (files, quizzes, workshops, ...)
  • TensorFlow

Sustainable Development Goals (SDGs)
Quality Education Industry, Innovation and Infrastructure